Deletion of Porcn in mice leads to multiple developmental defects and models human focal dermal hypoplasia (Goltz syndrome)

Abstract

Background: Focal Dermal Hypoplasia (FDH) is a genetic disorder characterized by developmental defects in skin, skeleton and ectodermal appendages. FDH is caused by dominant loss-of-function mutations in X-linked PORCN. PORCN orthologues in Drosophila and mice encode endoplasmic reticulum proteins required for secretion and function of Wnt proteins. Wnt proteins play important roles in embryo development, tissue homeostasis and stem cell maintenance. Since features of FDH overlap with those seen in mouse Wnt pathway mutants, FDH likely results from defective Wnt signaling but molecular mechanisms by which inactivation of PORCN affects Wnt signaling and manifestations of FDH remain to be elucidated.

Results: We introduced intronic loxP sites and a neomycin gene in the mouse Porcn locus for conditional inactivation. Porcn-ex3-7flox mice have no apparent developmental defects, but chimeric mice retaining the neomycin gene (Porcn-ex3-7Neo-flox) have limb, skin, and urogenital abnormalities. Conditional Porcn inactivation by EIIa-driven or Hprt-driven Cre recombinase results in increased early embryonic lethality. Mesenchyme-specific Prx-Cre-driven inactivation of Porcn produces FDH-like limb defects, while ectodermal Krt14-Cre-driven inactivation produces thin skin, alopecia, and abnormal dentition. Furthermore, cell-based assays confirm that human PORCN mutations reduce WNT3A secretion.

Conclusions: These data indicate that Porcn inactivation in the mouse produces a model for human FDH and that phenotypic features result from defective WNT signaling in ectodermal- and mesenchymal-derived structures.

Figure 1. Phenotype of chimeric Porcn genetrap mouse.

(A) Porcn genetrap construct CSD256. Top: A genetrap cassette with a strong splice acceptor (SA) upstream of an engrailed2 leader sequence (En2) fused to a β-galactosidase-neomycin (β-geo) gene is inserted in intron 2 of Porcn. Bottom: resulting genetrapped fusion transcript that truncates Porcn after exon 2 (green arrow, translation initiation site, A(n) poly-A tail). (B) Fused and hypoplastic shortened digits. (C) Areas of hypoplastic skin with alopecia (black arrows) and swelling at lower back. (D) Intra-abdominal cystic structure (likely dilated vas deferens), right hydronephrosis (arrow), testicle and seminiferous tubules (*). (E) Skeletal microCT images showing fusion and hypoplasia of digits in mutant mouse (WT, wild type mouse; CSD, mouse generated from CSD256; inset, enlargement of hindpaw). (F) Hypoplastic skin with dermoid cyst (*). Scale bar represents 1 mm. (G) Top panel: ovary and oviducts (scale bar represents 200 µm). Bottom panel: hypoplastic testis (scale bar represents 100 µm). (H) Top panel: hydronephrotic kidney with thin medulla. Bottom panel: contralateral normal kidney (scale bars represent 1 mm). (I) Gel picture showing amplification of lacZ (364 bp fragment) in various tissues. M, 1 kb ladder; 1, skin with lesion; 2, normal skin; 3, liver; 4, tail; P, positive control (CSD256 ES cell DNA); N, negative control (C57BL6 gDNA).

Figure 2. Phenotype of Porcn-ex3-7-Neo-flox chimeric mice.

(A) Hypoplastic, fused, and missing digits on right (R) and left (L) fore- (F) and hindlimbs (H) in different chimeras (C1–C4) compared to wild type (WT). (B) Skeletal preparations of extremities shown in panel A. (C) Vertebral abnormalities in the tail of chimera 1 (C1). (D) Hydronephrosis of the right kidney and normal left kidney in C1. (E) Hypoplastic testicle in C1 and normal testicle in C5. (F) Uterine abnormalities: asymmetrical hypoplastic uterine horn in C6. (G) Gel picture showing amplification of the targeted allele in various tissues of C1. H, heart; K(L), left kidney; LU, lung; SK, skin; SP, spleen; T, testis; K(R), right kidney; LI, liver; TG, targeted allele; WT, wild type allele.

Figure 3. Phenotype at E9.5 and gene expression analysis of X^{Porcn-ex3-7del}/X embryos.

(A) X^{Porcn-ex3-7del}/X embryo with open neural tube and abdominal wall closure defect (arrows). (B) Bar graph of quantitative real-time PCR analysis at E9.5 (top panel) and E10.5 (bottom panel) of Porcn wild type (n = 4 each) and mutant (n = 3 each) embryos. Bars are color-coded by gene type and indicate the log2 of the fold-deviation of gene expression levels in Porcn mutant embryos compared to wild type (set as 1). Error bars indicate standard errors of the mean. (*) indicates statistical significance at p<0.05 (Student's t-test).

Figure 4. Phenotype of X^{Porcn-Ex3-7flox}/Y;Prx-Cre mice.

(A) Shortened forelimbs at E18.5 in X^{Porcn-Ex3-7flox}/Y;Prx-Cre (PrxcKO) compared to wild type littermate (control). (B) Stunted growth at day 7 (P7) in PrxcKO compared to control littermate. (C) Appearance of PrxcKO compared to WT littermate at P28. (D) Digital abnormalities showing shortened limbs and shortened digits in PrxcKO compared to WT. F, forelimb; H, hindlimb. (E) Whole-skeleton preparations of newborn mice (NB). Mutant exhibits shortening of all skeletal elements. (F, G) Skeletal preparations of NB forelimbs and hindlimbs. (H) Whole-skeleton preparations of P28 mice. Mutant exhibits more dramatic shortening of all skeletal elements. (I, J) Skeletal preparations of forelimbs (F) and hindlimbs (H) at P28. hu, humerus; ra, radius; sc, scapula; ul, ulna; cl, clavicle; il, ilium; pu, pubis; fe, femur; ti, tibia; fi, fibula.

Figure 5. Phenotype of X^{Porcn-Ex3-7-flox}/Y;Krt14-Cre mice.

(A, B) Extensive alopecia of X^{Porcn-Ex3-7-flox}/Y;Krt14-Cre (Krt14cKO) compared to wild type (control) at P7. (C–E) Variable dental phenotype (missing and hypoplastic teeth) in Krt14cKO mice. (F–I) RNA in situ hybridization with a Porcn antisense (F, G) and sense (H, I) riboprobe to E16.5 Krt14cKO (G, I) and control (F, H) skin showing reduced Porcn expression in mutant. (J, K) H&E staining of dorsal skin of Krt14cKO (K) and control (J) mice at P9 shows lack of hair follicles in the Krt14cKO mice. (L, M) H&E staining of dorsal skin from newborn (NB) β-catenin ^flox/⁺;Krt14-Cre (β-catenin Krt14cKO; M) and control (L) mice also shows lack of hair follicles in the mutant. (N, O) H&E staining of skin from Krt14cKO (O) and control (N) mice at E16.5 shows that the early step of hair follicle morphogenesis does not take place in the Krt14cKO mice. hf, hair follicle. (P, Q) Immunostaining of skin at E16.5 shows that the Krt14cKO epidermis (Q) lacks Sox9-positive cells, which mark the hair follicle from the earliest budding stage (placode). (R, S) Skin of Krt14cKO epidermis (S) at E16.5 also lacks P-cadherin-positive cells, also indicating that placode formation is drastically compromised. Scale bars: F–I and L–S, 50 µm; J and K, 500 µm.

Figure 6. Oil red O staining of Krt14cKO skin at P9.

(A) Control skin. Scale bar 100 µm. (B–D) Krt14cKO skin shows fat apposition close to the surface. Scale bars 50 µm.

Figure 7. Cell-based WNT3A secretion assay.

(A) Western blot showing WNT3A levels in total cell lysates after transient transfection. (NTC, non-transfected control). (B) Quantification of WNT3A levels in cell lysates. Co-expression of WNT3A with wild type PORCN (WNT3A+PORCN) results in disappearance of WNT3A from the cell lysate compared to WNT3A alone or WNT3A+GFP. In contrast, co-expression of mutant PORCN forms p.M1I and p.R124X with WNT3A causes WNT3A retention in cells, but other mutations, p.S136F, p.G168R, and p.Y359X, do not affect WNT3A secretion. We compared WNT3A levels in cells co-transfected by WNT3A and either wild type or mutant PORCN forms to WNT3A levels in cells expressing WNT3A+GFP, because these contain similar amounts of transfected DNA. Fold changes with standard deviation are shown; all data were normalized to β-tubulin; (*) indicates significant difference at p<0.05; the experiment was repeated 4 times.